Confusion Matrix
A confusion matrix presents classifier predictions versus true labels as TP/FP/TN/FN counts in matrix form; it is the foundation of classification evaluation.
Purpose
To evaluate per-class classifier performance comprehensively and compute precision/recall/F1.
Mechanism
Rows are true classes and columns are predicted classes; the diagonal holds correct classifications and off-diagonal cells hold errors. Binary case: TP (true positive), FP (false positive), TN (true negative), FN (false negative). Multiclass extends to an N×N matrix.
Use cases
- Classification evaluation
- Metric computation
- Model diagnosis